Improve model card for Variational Reasoning for Language Models

README.md

@@ -1,198 +1,86 @@
 ---
 library_name: transformers
-tags: []
+tags:
+- reasoning
+- qwen
+pipeline_tag: text-generation
+language: en
 ---
 
-# Model Card for Model ID
-
-<!-- Provide a quick summary of what the model is/does. -->
-
+# Model Card for Variational Reasoning for Language Models
 
+This repository contains the models presented in the paper [Variational Reasoning for Language Models](https://huggingface.co/papers/2509.22637).
 
 ## Model Details
 
 ### Model Description
 
-<!-- Provide a longer summary of what this model is. -->
-
-This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
+We introduce a variational reasoning framework for language models that treats thinking traces as latent variables and optimizes them through variational inference. Starting from the evidence lower bound (ELBO), we extend it to a multi-trace objective for tighter bounds and propose a forward-KL formulation that stabilizes the training of the variational posterior. We further show that rejection sampling finetuning and binary-reward RL, including GRPO, can be interpreted as local forward-KL objectives, where an implicit weighting by model accuracy naturally arises from the derivation and reveals a previously unnoticed bias toward easier questions. We empirically validate our method on the Qwen 2.5 and Qwen 3 model families across a wide range of reasoning tasks. Overall, our work provides a principled probabilistic perspective that unifies variational inference with RL-style methods and yields stable objectives for improving the reasoning ability of language models.
 
-- **Developed by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Model type:** [More Information Needed]
-- **Language(s) (NLP):** [More Information Needed]
-- **License:** [More Information Needed]
-- **Finetuned from model [optional]:** [More Information Needed]
+-   **Developed by:** Xiangxin Zhou, Zichen Liu, Haonan Wang, Chao Du, Min Lin, Chongxuan Li, Liang Wang, Tianyu Pang
+-   **Model type:** Causal Language Model
+-   **Language(s) (NLP):** English
+-   **License:** [More Information Needed]
+-   **Finetuned from model:** Qwen3-4B-Base, Qwen3-8B-Base, Qwen2.5-7B-Instruct, Qwen2.5-32B-Instruct (as described in the GitHub repository's "Models and Datasets" table, serving as backbones for the variational reasoning framework)
 
-### Model Sources [optional]
+### Model Sources
 
-<!-- Provide the basic links for the model. -->
-
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
+-   **Repository:** https://github.com/sail-sg/variational-reasoning
+-   **Paper:** https://huggingface.co/papers/2509.22637
 
 ## Uses
 
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
-
 ### Direct Use
 
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
-
-[More Information Needed]
-
-### Downstream Use [optional]
-
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
-
-[More Information Needed]
+This model is intended for research and development focused on improving the reasoning capabilities of language models. It can be used for tasks requiring complex, multi-step thinking, leveraging the variational inference framework.
 
 ### Out-of-Scope Use
 
-<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
-
-[More Information Needed]
-
-## Bias, Risks, and Limitations
-
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
-
-[More Information Needed]
-
-### Recommendations
-
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
-
-Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+This model should not be used for generating harmful, biased, or unethical content. Users should exercise caution and ensure responsible deployment, especially in sensitive applications, as comprehensive safety evaluations are beyond the scope of this research paper.
 
 ## How to Get Started with the Model
 
-Use the code below to get started with the model.
-
-[More Information Needed]
+For detailed usage instructions, including evaluation and training pipelines, please refer to the [official GitHub repository](https://github.com/sail-sg/variational-reasoning). The repository provides scripts and guidelines to get started with the data processing, training, and evaluation suite.
 
 ## Training Details
 
 ### Training Data
 
-<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
-
-[More Information Needed]
+The project utilizes various datasets for training the different components of the variational reasoning framework, such as `Variational-Posterior-4B-Acc-mix`, `Variational-Posterior-4B-GML-mix`, etc., as listed in the GitHub repository. Detailed information on data processing and dataset specifics can be found in the [official GitHub repository](https://github.com/sail-sg/variational-reasoning).
 
 ### Training Procedure
 
-<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
-
-#### Preprocessing [optional]
-
-[More Information Needed]
-
+The models are trained using a multi-stage process involving an initial reasoning model ($\pi_{\theta_0}$), a variational posterior ($q_\phi$), and a final reasoning model ($\pi_\theta$). The training pipelines are initialized from [LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory) and `SkyThought` is used for verification and evaluation. Detailed scripts and configuration files can be found within the `LLaMA-Factory/variational_reasoning/train` directory of the GitHub repository.
 
 #### Training Hyperparameters
 
-- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
-
-#### Speeds, Sizes, Times [optional]
-
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
-
-[More Information Needed]
+Training scripts assume 2 nodes (2 x 8 H100 GPUs), with `gradient_accumulation_steps` adjusted accordingly for different setups. Specific hyperparameters are detailed in `LLaMA-Factory/examples/variational_reasoning/*.yaml` files in the GitHub repository.
 
 ## Evaluation
 
-<!-- This section describes the evaluation protocols and provides the results. -->
-
-### Testing Data, Factors & Metrics
-
-#### Testing Data
-
-<!-- This should link to a Dataset Card if possible. -->
-
-[More Information Needed]
-
-#### Factors
-
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
-
-[More Information Needed]
-
-#### Metrics
-
-<!-- These are the evaluation metrics being used, ideally with a description of why. -->
-
-[More Information Needed]
-
-### Results
-
-[More Information Needed]
-
-#### Summary
-
-
-
-## Model Examination [optional]
-
-<!-- Relevant interpretability work for the model goes here -->
-
-[More Information Needed]
+The evaluation of the models is conducted using an evaluation suite, and instructions can be found in `SkyThought/variational_reasoning/eval/eval.sh` within the [official GitHub repository](https://github.com/sail-sg/variational-reasoning). Models have been empirically validated on the Qwen 2.5 and Qwen 3 model families across a wide range of reasoning tasks.
 
 ## Environmental Impact
 
-<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
-
 Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
 
-- **Hardware Type:** [More Information Needed]
-- **Hours used:** [More Information Needed]
-- **Cloud Provider:** [More Information Needed]
-- **Compute Region:** [More Information Needed]
-- **Carbon Emitted:** [More Information Needed]
-
-## Technical Specifications [optional]
-
-### Model Architecture and Objective
-
-[More Information Needed]
-
-### Compute Infrastructure
-
-[More Information Needed]
-
-#### Hardware
-
-[More Information Needed]
-
-#### Software
-
-[More Information Needed]
-
-## Citation [optional]
-
-<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
-
-**BibTeX:**
-
-[More Information Needed]
-
-**APA:**
-
-[More Information Needed]
-
-## Glossary [optional]
-
-<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
-
-[More Information Needed]
-
-## More Information [optional]
-
-[More Information Needed]
-
-## Model Card Authors [optional]
-
-[More Information Needed]
+-   **Hardware Type:** [More Information Needed]
+-   **Hours used:** [More Information Needed]
+-   **Cloud Provider:** [More Information Needed]
+-   **Compute Region:** [More Information Needed]
+-   **Carbon Emitted:** [More Information Needed]
+
+## Citation
+
+If you find this work useful, please consider citing our paper:
+```bib
+@article{zhou2025variationalreasoninglanguagemodels,
+      title={Variational Reasoning for Language Models}, 
+      author={Xiangxin Zhou and Zichen Liu and Haonan Wang and Chao Du and Min Lin and Chongxuan Li and Liang Wang and Tianyu Pang},
+      journal={arXiv preprint arXiv:2509.22637},
+      year={2025}
+}
+```
 
 ## Model Card Contact